Satellite-borne sar image on-orbit radiometric and geometric correction method based on parallel computing architecture

Abstract

The application relates to the technical field of synthetic aperture radar signal processing, and discloses a spaceborne SAR image on-orbit radiation and geometric correction method based on a parallel computing architecture, the radiation correction method comprising the following steps: acquiring ephemeris parameters of a satellite in a geocentric coordinate system at each azimuth time; calculating the slant range between the satellite beam center pointing to the scene center point and the antenna phase center at each azimuth time and the slant range between each sampling point in the SAR echo distance direction and the satellite at each azimuth time; calculating the downward angle between the satellite beam center pointing at each azimuth time and the satellite and the subsatellite point connecting line; calculating the antenna distance direction beam pattern and constructing a radiation correction function; and performing parallel correction on the radiation distortion of the SAR image based on the radiation correction function. The application effectively improves the SAR image quality and accurate positioning of the SAR image, meets the radiation fidelity of the SAR image, improves the data processing task efficiency, and meets the task demand of high real-time requirement.

Description

Technical Field

[0001] This invention relates to the field of synthetic aperture radar signal processing technology, and in particular to an on-orbit radiometric and geometric correction method for spaceborne SAR images based on a parallel computing architecture. Background Technology

[0002] Synthetic Aperture Radar (SAR), as an active, high-resolution microwave imaging system, can continuously observe large areas of interest. In recent years, SAR has evolved towards multi-polarization and multi-imaging modes, playing an important role in ecological environment surveying, military reconnaissance, and civilian applications. This requires SAR images to not only have high resolution but also high radiometric and geometric fidelity to further improve image quality and better meet a wide range of application needs. The widespread availability of high-quality, high-resolution SAR images has also promoted the real-time demand for rapid extraction of remote sensing image information. However, the response time of traditional remote sensing systems based on satellite data transmission to ground stations for processing is too long, making it difficult to meet the high real-time requirements of various applications. Radiometric correction and geometric correction are two key steps in the on-orbit preprocessing of spaceborne SAR images. Radiometric correction corrects the radiometric distortion of the entire SAR image scene, improving SAR image quality; geometric correction enables the correct correlation and effective recovery of geographic coordinates and original pixel information, achieving accurate positioning of target information in SAR images. Therefore, how to perform rapid radiometric and geometric correction on the entire SAR image scene without radiometric calibration data and digital elevation models is an application requirement that urgently needs to be addressed to improve the real-time performance of on-orbit SAR image preprocessing, and it has important application value for the development of the national economy and national defense. Summary of the Invention

[0003] The technical problem solved by this invention is to provide an on-orbit radiometric and geometric correction method for spaceborne SAR images based on a parallel computing architecture, which effectively solves the technical problems of low SAR image quality and slow processing efficiency.

[0004] The above-mentioned technical problems are solved by the following technical solutions:

[0005] An on-orbit radiometric correction method for spaceborne SAR images based on a parallel computing architecture, the method comprising:

[0006] Obtain the ephemeris parameters of the satellite in the geocentric coordinate system at various positions and times. The ephemeris parameters include at least one of the following: position vector, velocity vector, three-axis attitude angle, pulse repetition frequency, wave position sampling start time, signal pulse width, signal wavelength, antenna range aperture length, antenna azimuth aperture length, antenna beam elevation angle, antenna beam azimuth angle, number of range sampling points, number of azimuth sampling points, and signal sampling frequency.

[0007] Calculate the slant distance between the satellite beam center pointing to the scene center point and the antenna phase center at each azimuth time, and the slant distance between each sampling point and the satellite at each azimuth time as well as the SAR echo distance at each azimuth time.

[0008] Based on the slant distance between the satellite beam center pointing to the scene center point and the antenna phase center at each azimuth time, the slant distance between the SAR echo distance at each azimuth time and each sampling point and the satellite, and the ephemeris parameters, the downward angle between the satellite beam center pointing to each azimuth time and the line connecting the satellite and the nadir point is calculated.

[0009] Based on the satellite beam center pointing at each of the aforementioned time points and the downward angle between the satellite and the line connecting the nadir point, the antenna range-beam pattern is calculated and a radiation correction function is constructed.

[0010] The radiation distortion of SAR images is corrected in parallel based on the aforementioned radiation correction function.

[0011] Beneficial effects: By acquiring the ephemeris parameters of the satellite in the geocentric coordinate system at each azimuth time, the slant range between the satellite beam center pointing to the scene center point and the antenna phase center at each azimuth time, as well as the slant range between the SAR echo range and each sampling point and the satellite at each azimuth time, are calculated. Furthermore, based on the slant range and ephemeris parameters, the downward angle between the satellite beam center pointing and the line connecting the satellite and the nadir point at each azimuth time is calculated. Then, based on the downward angle between the satellite beam center pointing and the line connecting the satellite and the nadir point at each azimuth time, the antenna range-beam pattern is calculated, and a radiometric correction function is constructed. Finally, the radiometric distortion of the SAR image is corrected in parallel using the radiometric correction function. This effectively improves the SAR image quality, meets the radiometric fidelity requirement of SAR images, improves the efficiency of data processing tasks, and also meets the requirements of tasks with high real-time performance.

[0012] In one embodiment, the parallel correction of radiometric distortion of the SAR image based on the radiometric correction function includes:

[0013] The SAR image is divided into multiple sub-SAR images;

[0014] Based on the CUDA programming model, thread blocks are set up corresponding to the multiple sub-SAR images;

[0015] The parallel processing of multiple thread blocks using a graphics processor completes the parallel correction of radiometric distortion in multiple sub-SAR images.

[0016] In one embodiment, the three-axis attitude angles include the ground roll angle, the ground pitch angle, and the ground yaw angle; the calculation of the slant range between the satellite beam center pointing to the scene center point and the antenna phase center at each azimuth time and the slant range between the SAR echo range at each sampling point and the satellite at each azimuth time based on the ephemeris parameters includes:

[0017] Based on the three-axis attitude angles and the satellite's velocity and position vectors in the geocentric coordinate system, the number of nadir echo pulses and the SAR echo acquisition start time are calculated. The calculation formula is shown in equation (1):

[0018]

[0019] Based on the start time of the echo acquisition and the number of echo pulses at the nadir point, the slant distance between the satellite at the beam center and the beam pointing point, and the slant distance between each sampling point and the satellite at each azimuth time SAR echo distance are calculated. The calculation formula is shown in Equation (2):

[0020]

[0021] Where c is the speed of light; R j R represents the slant range between the SAR echo range at each sampling point and the satellite at each azimuth time. ref The slant distance between the satellite beam center pointing to the scene center point and the antenna phase center at each location and time; t center t0 is the start time of echo acquisition; M is the number of echo pulses at the nadir point; t0 is the start time of waveform sampling; PRF is the pulse repetition frequency; T p f is the signal pulse width; s N is the signal sampling frequency; r This represents the number of sampling points in the distance direction.

[0022] In one embodiment, the calculation of the downward angle between the satellite beam center pointing to the scene center point and the antenna phase center based on the slant range of the satellite beam center pointing to the scene center point and the antenna phase center at each azimuth time, the slant range of the SAR echo distance at each azimuth time to each sampling point and the satellite, and the ephemeris parameters includes:

[0023] The distance between the satellite and the nadir point is calculated based on the satellite's position vector and the Earth's average radius.

[0024] Based on the slant distance between the satellite beam center pointing to the scene center point and the antenna phase center at each azimuth time, the slant distance between the SAR echo distance at each azimuth time and each sampling point and the satellite, the distance between the satellite and the nadir point, and the average radius of the Earth, the downward angle between the satellite beam center pointing to the scene center point and the line connecting the satellite and the nadir point at each azimuth time is calculated; the calculation formula is shown in equation (3):

[0025]

[0026] Where, β j R represents the downward angle between the satellite beam center pointing at each location and the line connecting the satellite and the nadir point; H represents the distance from the satellite to the nadir point; R represents the downward angle between the satellite beam center pointing at each location and the line connecting the satellite and the nadir point. e This is the average radius of the Earth.

[0027] In one embodiment, the step of calculating the antenna range-direction beam pattern and constructing a radiation correction function based on the satellite beam center pointing at the stated locations and the downward viewing angle between the satellite and the nadir point includes:

[0028] The difference between the downward angle of the satellite beam center and the line connecting the satellite and the nadir point at each time and location and the downward angle corresponding to the center of the scene is calculated using the formula shown in equation (4):

[0029] Δβ=β j -β ref (4)

[0030] The antenna range-direction beam pattern is calculated based on the difference Δβ, and the calculation formula is shown in equation (5):

[0031]

[0032] The radiation correction function is constructed based on the antenna range-direction beam pattern, and the calculation formula is shown in equation (6):

[0033]

[0034] Where, β j The downward angle between the satellite beam center pointing at each position and the line connecting the satellite and the nadir point; β ref Δβ represents the downward viewing angle corresponding to the scene center; Δβ is the difference between the downward viewing angle corresponding to each sampling point in the SAR echo range at each azimuth time and the downward viewing angle corresponding to the scene center; W a (β j H) represents the antenna range-to-beam pattern; L r λ is the antenna distance-direction aperture length; λ is the signal wavelength; F(W) a (β jH)) is the radiometric correction function; α is an adjustable scaling factor after image normalization; n is a constant, n∈[0,4].

[0035] On the other hand, the present invention also provides an on-orbit geometric correction method for spaceborne SAR images based on a parallel computing architecture, the method further comprising:

[0036] Based on the satellite's position vector, velocity vector, and antenna parameters in the satellite's geocentric coordinate system, calculate the coordinates of the scene point illuminated by the antenna beam and the Doppler center frequency of the satellite relative to the antenna beam aiming point.

[0037] The average elevation information of the target scene is obtained, and the Doppler center frequency is combined in the range Doppler SAR positioning model to generate the coordinates and latitude and longitude information of each pixel in the SAR image in the geocentric coordinate system.

[0038] Establish an index mapping relationship between the coordinates of each pixel in the SAR image in the geocentric coordinate system and the geographic latitude and longitude grid;

[0039] A kernel function based on a graphics processor is constructed to sequentially map SAR image pixels in the imaging plane to a latitude and longitude grid, thus completing geometric correction.

[0040] In one embodiment, the calculation of the coordinates of the ground scene point illuminated by the antenna beam and the Doppler center frequency of the satellite relative to the antenna beam aiming point, based on the satellite's position vector, velocity vector, and antenna parameters in the satellite's geocentric coordinate system, includes:

[0041] Calculate the rotation matrix from the inertial coordinate system to the satellite orbital coordinate system based on the satellite's ephemeris parameters;

[0042] Based on the satellite's three-axis attitude angles and the satellite's rotation sequence, calculate the rotation matrix from the satellite's orbital coordinate system to the satellite's body coordinate system;

[0043] Calculate the rotation matrix from the satellite coordinate system to the antenna coordinate system based on the antenna azimuth and elevation angles in the antenna coordinate system.

[0044] Calculate the intersection of the aiming vector of the antenna beam center with the Earth's surface in the geocentric coordinate system to obtain the coordinates of the scene point illuminated by the antenna beam on the ground, and calculate the velocity vector of the antenna beam in the geocentric coordinate system.

[0045] Based on the coordinates of the scene point illuminated by the antenna beam and the velocity vector of the antenna beam in the geocentric coordinate system, the Doppler center frequency of the satellite relative to the antenna beam aiming point is calculated.

[0046] In one embodiment, the step of acquiring the average elevation information of the target scene and combining it with the Doppler center frequency in the range-Doppler SAR positioning model to generate the geocentric coordinates and latitude and longitude information of each pixel in the SAR image includes:

[0047] Based on the average elevation information of the target scene and the coordinates and latitude and longitude information of each pixel in the geocentric coordinate system in the SAR image;

[0048] In the range-Doppler SAR positioning model, based on the average elevation information and Doppler center frequency of the target scene, the coordinates of the four corner points in the SAR image in the geocentric coordinate system are calculated, and the coordinates are converted into latitude and longitude information; the longitude Ω of the four corner points are respectively {Ω left_up Ω left_down Ω right_up Ω right_down}, the latitudes Φ are respectively {Φ left_up Φ left_down , Φ right_up Φ right_down};

[0049] Compare the top left corner point Ω left_up and the latitude value Ω of the bottom left corner point left_down Determine whether the scene pixels are consistent with the direction from north latitude to south latitude;

[0050] In response to the fact that the scene pixels are aligned with the direction from north to south latitude, the latitude and longitude values ​​corresponding to all pixels in the SAR image are calculated based on the latitude and longitude of the four corner points and by constructing a kernel function.

[0051] On the other hand, the present invention also provides an electronic device, including: a memory, a processor, a communication interface and a communication bus, wherein the processor, the memory and the communication interface communicate with each other through the communication bus;

[0052] The memory is used to store at least one executable instruction that causes the processor to execute the on-orbit radiometric and geometric correction method for spaceborne SAR images based on a parallel computing architecture as described above.

[0053] On the other hand, the present invention also provides a computer-readable storage medium storing computer instructions for causing a processor to execute the on-orbit radiometric and geometric correction method for spaceborne SAR images based on a parallel computing architecture as described above. Attached Figure Description

[0054] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0055] Figure 1 This is a flowchart illustrating an on-orbit radiometric correction method for spaceborne SAR images based on a parallel computing architecture, according to an embodiment of the present invention.

[0056] Figure 2 This is a flowchart illustrating another embodiment of the present invention: an on-orbit radiometric correction method for spaceborne SAR images based on a parallel computing architecture.

[0057] Figure 3 This is a flowchart illustrating an on-orbit geometric correction method for spaceborne SAR images based on a parallel computing architecture, according to an embodiment of the present invention.

[0058] Figure 4 This is a flowchart illustrating an on-orbit radiometric and geometric correction method for spaceborne SAR images based on a parallel computing architecture, according to an embodiment of the present invention.

[0059] Figure 5 A SAR image of a certain scene without radiometric correction;

[0060] Figure 6 The image is a SAR image after radiometric correction using an on-orbit radiometric correction method for spaceborne SAR images based on a parallel computing architecture, according to an embodiment of the present invention.

[0061] Figure 7 The original SAR image of a certain scene;

[0062] Figure 8 The image is a SAR image corrected by an on-orbit radiometric and geometric correction method for spaceborne SAR images based on a parallel computing architecture according to an embodiment of the present invention.

[0063] Figure 9 This is a schematic diagram of the structure of an electronic device according to an embodiment of the present invention. Detailed Implementation

[0064] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0065] In the description of this invention, it should be understood that the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this invention, unless otherwise stated, "a plurality of" means two or more. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0066] According to embodiments of the present invention, such as Figure 1 , Figure 2 and Figure 4 As shown, an on-orbit radiometric correction method for spaceborne SAR images based on a parallel computing architecture is provided, including the following steps:

[0067] Step S100: Obtain the ephemeris parameters of the satellite in the geocentric coordinate system at various positions and times. The ephemeris parameters include the position vector P. sat =[X sat Y sat Z sat ], velocity vector V sat =[Vx sat Vx sat Vz sat [ ] Three-axis attitude angle, pulse repetition frequency (PRF), wave position sampling start time (t0), signal pulse width (T p Signal wavelength λ, antenna distance to aperture length L r Antenna azimuth aperture length L a Antenna beam elevation angle Antenna beam azimuth Number of sampling points N in the distance direction r Number of azimuth sampling points N a and signal sampling frequency f s etc.; Three-axis attitude angles include the roll angle θ relative to the ground. Roll Pitch angle θ Pitch and ground yaw angle θ Yaw .

[0068] Step S200: Calculate the slant range R between the satellite beam center pointing to the scene center point and the antenna phase center at each location and time. ref And the slant range R between each sampling point and the satellite at each azimuth time point of the SAR echo range. j ,j∈[1:N r ].

[0069] In step S200, specifically, the number of nadir echo pulses and the start time of SAR echo acquisition are calculated based on the three-axis attitude angles and the satellite's velocity and position vectors in the geocentric coordinate system. The calculation formula is shown in equation (1):

[0070]

[0071] Based on the start time of echo acquisition and the number of echo pulses at the nadir point, the slant distance between the satellite at the beam center and the beam pointing point, and the slant distance between each sampling point and the satellite at each azimuth time of the SAR echo distance are calculated. The calculation formula is shown in Equation (2):

[0072]

[0073] Where c is the speed of light; R is the slant distance between each sampling point and the satellite at each azimuth time; R ref The slant distance between the satellite beam center pointing to the scene center point and the antenna phase center at each location and time; t center t0 is the start time of echo acquisition; M is the number of echo pulses at the nadir point; t0 is the start time of waveform sampling; PRF is the pulse repetition frequency; T p f is the signal pulse width; s N is the signal sampling frequency; r This represents the number of sampling points in the distance direction.

[0074] Step S300: Based on the slant range between the satellite beam center pointing to the scene center point and the antenna phase center at each azimuth time, the slant range between the SAR echo distance to each sampling point and the satellite at each azimuth time, and the ephemeris parameters, calculate the downward angle β between the satellite beam center pointing to the scene center point and the line connecting the satellite and the nadir point at each azimuth time. j , j∈[1:N r ].

[0075] Step S300 specifically includes calculating the distance between the satellite and the nadir point based on the satellite's position vector and the Earth's average radius;

[0076] Based on the slant distance between the satellite beam center pointing to the scene center point and the antenna phase center at each azimuth time, the slant distance between each sampling point and the satellite at each azimuth time, the distance from the satellite to the nadir point, and the average radius of the Earth, the downward angle between the satellite beam center pointing to the scene center point and the line connecting the satellite and the nadir point at each azimuth time is calculated; the calculation formula is shown in equation (3):

[0077]

[0078] Where βi is the downward angle between the satellite beam center pointing at each location and the line connecting the satellite and the nadir point; H is the distance from the satellite to the nadir point; Re This is the average radius of the Earth.

[0079] The formula for calculating the distance H between the satellite and the nadir point based on the satellite's position vector and the Earth's average radius is as follows:

[0080] H = sqrt(X) sat 2 +Y sat 2 +Z sat 2 )-R e .

[0081] Step S400: Calculate the antenna range-to-beam pattern W based on the satellite beam center pointing at each location and the downward viewing angle between the satellite and the line connecting the nadir point. a (β, H) and construct the radiation correction function F(W) a (β j H)).

[0082] In step S400, the specific steps include calculating the difference between the downward angle of the satellite beam center pointing and the line connecting the satellite and the nadir point at each location and the downward angle corresponding to the scene center. The calculation formula is shown in equation (4):

[0083] Δβ=β j -β ref (4)

[0084] The antenna range-direction beam pattern is calculated based on the difference Δβ, and the calculation formula is shown in equation (5):

[0085]

[0086] The radiation correction function is constructed based on the antenna range-direction beam pattern, and the calculation formula is shown in equation (6):

[0087]

[0088] Where, β j The downward angle between the satellite beam center pointing at each position and the line connecting the satellite and the nadir point; β ref Δβ represents the downward viewing angle corresponding to the scene center; Δβ is the difference between the downward viewing angle corresponding to each sampling point in the SAR echo range at each azimuth time and the downward viewing angle corresponding to the scene center; W a (β j H) represents the antenna range-to-beam pattern; L r λ is the antenna distance-direction aperture length; λ is the signal wavelength; F(W) a (β jH)) is the radiometric correction function; α is an adjustable scaling factor after image normalization; n is a constant, n∈[0,4].

[0089] Step S500: Based on the radiation correction function F(W) a (β j H)) corrects the radiation distortion of SAR images.

[0090] Step S500 also includes the following steps:

[0091] Step S510: Divide the SAR image into multiple sub-SAR images;

[0092] Step S520: Set up thread blocks corresponding to multiple sub-SAR images based on the CUDA programming model;

[0093] Step S530: Perform parallel computation on multiple thread blocks based on the graphics processor to complete the parallel correction of radiation distortion of multiple sub-SAR images.

[0094] Specifically, first, the SAR image is copied from the host to the device; then, a file of size N is transferred to the device. a ×N r The SAR image is divided into n blocks, resulting in n sub-SAR image blocks. Furthermore, based on a parallel computing platform and programming model (Compute Unified Device Architecture, CUDA), the size of the two-dimensional grid corresponding to the n sub-SAR images is set to grid = {G}. a G r There are} thread blocks, each two-dimensional thread block having a size of block = {B}. a B r Therefore, the number of pixels that each thread block can process is:

[0095]

[0096] Then, a graphics processing unit (GPU) is used to perform multi-threaded parallel processing on the divided multiple thread blocks, and each pixel unit is compared with the correction function F(W). a (β j The coefficients of H)) are multiplied and processed in parallel using multi-threaded computation to achieve parallel correction of radiometric distortion in SAR images at the block level, thus completing the radiometric correction of SAR images. Parallel correction of radiometric distortion in SAR images using GPUs can effectively improve the correction speed to meet the demands of real-time tasks.

[0097] In this embodiment, by acquiring the ephemeris parameters of the satellite in the geocentric coordinate system at each azimuth time, the slant range between the satellite beam center pointing to the scene center point and the antenna phase center at each azimuth time, as well as the slant range between each sampling point of the SAR echo range and the satellite at each azimuth time, are calculated. Furthermore, based on the slant range and ephemeris parameters, the downward angle between the satellite beam center pointing and the line connecting the satellite and the nadir point at each azimuth time is calculated. Then, based on the downward angle between the satellite beam center pointing and the line connecting the satellite and the nadir point at each azimuth time, the antenna range-beam pattern is calculated, and a radiometric correction function is constructed. Finally, the radiometric distortion of the SAR image is corrected in parallel using the radiometric correction function. This effectively improves the SAR image quality, meets the radiometric fidelity requirement of SAR images, improves the efficiency of data processing tasks, and also meets the requirements of tasks with high real-time performance.

[0098] like Figure 5 and Figure 6 As shown, Figure 5 The imaging results are for SAR images that have not undergone radiometric correction. Figure 6 The images shown are SAR images corrected using the on-orbit radiometric correction method for spaceborne SAR images based on a parallel computing architecture provided in this embodiment of the invention. The comparison shows that the radiometrically corrected SAR images have relatively high clarity, balanced overall brightness, and a significant improvement in overall SAR image quality.

[0099] According to an embodiment of the present invention, on the other hand, such as Figure 3 , Figure 4 As shown, an on-orbit geometric correction method for spaceborne SAR images based on a parallel computing architecture is provided, including the following steps:

[0100] Step S600: Calculate the coordinates of the ground scene point illuminated by the antenna beam and the Doppler center frequency of the satellite relative to the antenna beam aiming point based on the satellite's position vector, velocity vector and antenna parameters in the satellite's geocentric coordinate system.

[0101] Step S600 specifically includes:

[0102] Calculate the rotation matrix from the inertial coordinate system to the satellite orbital coordinate system based on the satellite's ephemeris parameters;

[0103] Based on the satellite's three-axis attitude angles and the satellite's rotation sequence, calculate the rotation matrix from the satellite's orbital coordinate system to the satellite's body coordinate system;

[0104] Calculate the rotation matrix from the satellite coordinate system to the antenna coordinate system based on the antenna azimuth and elevation angles in the antenna coordinate system.

[0105] Calculate the intersection of the aiming vector of the antenna beam center with the Earth's surface in the geocentric coordinate system to obtain the coordinates of the scene point illuminated by the antenna beam on the ground, and calculate the velocity vector of the antenna beam in the geocentric coordinate system.

[0106] Based on the coordinates of the scene point illuminated by the antenna beam and the velocity vector of the antenna beam in the geocentric coordinate system, the Doppler center frequency of the satellite relative to the antenna beam aiming point is calculated.

[0107] Specifically, based on the satellite's position vector P in the geocentric coordinate system sat =[X sat Y sat Z sat ], velocity vector V sat =[Vx sat Vx sat Vz sat ] Calculate the rotation matrix from the inertial coordinate system to the satellite orbital coordinate system. Due to the calculation of velocity vector V sat_ω We need to take Earth's rotation into account, then V sat_ω =V sat +ω e ×P sat ; where ω e The angular velocity of Earth's rotation, i.e., ω e = [0, 0, 7.3e-5].

[0108] Based on the roll angle θ Roll Pitch angle θ pitch yaw angle θ Yaw And the rotation sequence of the satellite is defined to calculate the rotation matrix A from the satellite orbital coordinate system to the satellite body coordinate system. ve Further based on the antenna beam elevation angle Antenna beam azimuth Calculate the rotation matrix A from the satellite coordinate system to the antenna coordinate system. ea Assuming the aiming line of the antenna beam center coincides with the Z-axis of the antenna coordinate system, the aiming vector P of the antenna beam center can be obtained. e =[X e Y e Z e The expression in the central coordinate system:

[0109] P e =A ov ·A ve ·A ea ·[0, 0, 1] T

[0110] Based on the slant range R between the satellite beam center pointing to the scene center point and the antenna phase center...ref And the aiming vector P of the antenna beam center e =[X e Y e Z e The coordinates P of the scene point illuminated by the antenna beam were calculated. tar =[X tar Y tar Z tar ], and calculate the velocity vector V of the antenna beam in the geocentric coordinate system. tar =[Vx tar Vy tar Vz tar The calculation formula is:

[0111]

[0112] Then, based on the coordinates of the scene point illuminated by the antenna beam and the velocity vector of the antenna beam in the geocentric coordinate system, the Doppler center frequency of the satellite relative to the antenna beam aiming point is calculated. The calculation formula is shown in equation (7):

[0113]

[0114] Among them, f d P is the Doppler center frequency; tar P represents the coordinates of the point on the ground illuminated by the antenna beam. sat V is the satellite's position vector; sat V is the satellite's velocity vector. tar λ is the velocity vector of the antenna beam in the geocentric coordinate system; λ is the signal wavelength.

[0115] Furthermore, to calculate the distance R between the intersection point of the centroid of the antenna coordinate system and the aiming line of the antenna beam center with the Earth's surface... min Number of echo pulses M at the sub-satellite point b , will P e =[X e Y e Z e ] and P sat =[X sat Y sat Z sat Substituting into the ellipsoid equation, we can solve for R. min and M b :

[0116]

[0117] Where a is the Earth's major axis and b is the Earth's minor axis.

[0118] Step S700: Obtain the average elevation information of the target scene, and combine it with the Doppler center frequency in the range-Doppler SAR positioning model to generate the coordinates and latitude and longitude information of each pixel in the geocentric coordinate system.

[0119] Step S700 specifically includes the average elevation information of the target scene and the coordinates and latitude and longitude information of each pixel in the SAR image in the geocentric coordinate system.

[0120] In the range-Doppler SAR positioning model, based on the average elevation information and Doppler center frequency of the target scene, the coordinates of the four corner points in the SAR image in the geocentric coordinate system are calculated, and the coordinates are converted into latitude and longitude information; the longitude Ω of the four corner points are respectively {Ω left_up Ω left_down Ω right_up Ω right_down}, the latitudes Φ are respectively {Φ left_up Φ left_down Φ right_up Φ right_down};

[0121] Compare the top left corner point Ω left_up and the latitude value Ω of the bottom left corner point left_down Determine whether the scene pixels are consistent with the direction from north latitude to south latitude;

[0122] In response to the alignment of scene pixels with latitude from north to south, the latitude and longitude values ​​of all pixels in the SAR image are calculated based on the latitude and longitude of the four corner points and by constructing a kernel function.

[0123] Specifically, to more clearly describe the relationship between SAR image coordinates (i,j) and latitude / longitude grid coordinates (m,n), the index of the azimuth row number i and the index of the range column number j of the SAR image are first converted into the slant range R and the azimuth time t. a It can be represented as shown in the following formula:

[0124]

[0125] Then, based on the Earth model, slant range history, and Doppler equations, a range-Doppler positioning equation is established to calculate the position coordinates of each pixel in the SAR image in the geocentric coordinate system. The expression is:

[0126]

[0127] Where h is the average elevation of the target scene, (X T ,Y T Z T ) represents the geographic coordinates of each pixel in the SAR image to be solved.

[0128] Further, Newton's iterative method is used to solve (X). T ,Y T Z T ),set up The iterative formula for solving the range-Doppler positioning equation can then be expressed by the following generalized operator:

[0129]

[0130] in, Let be the inverse matrix of the solution at the k-th iteration, then:

[0131]

[0132] In each iterative solution (X) T ,Y T Z T To accelerate convergence, the center point P of each row of pixels in the SAR image is selected. tar =[X tar ,Y tar Z tar The coordinates of [ ] are used as the initial value. The value T is calculated after each iteration. k+1 and T k The maximum difference in the x-axis, y-axis, and z-axis directions, i.e.:

[0133]

[0134] When σ k+1 When the value is less than 1e-7, stop the iteration and output (X). T ,Y T Z T The solution results are as follows.

[0135] The coordinates (X, J) corresponding to each pixel (i, j) in the SAR image obtained by solving T ,Y T Z T Substituting these values ​​into the following formula, we can then calculate the longitude Φ corresponding to the pixel (i,j). The calculation expression is as follows:

[0136]

[0137] The latitude Ω corresponding to pixel (i,j) requires k iterations to calculate until the latitude Ω value converges. The parameters needed to calculate the latitude value include the Earth's radius R. e The extreme flattening of the reference ellipsoid and the radius of curvature N of the reference ellipsoid e The formula for iteratively calculating latitude is shown below:

[0138]

[0139] Combining the average elevation information of the target scene with the coordinates of the four corner points of the SAR image in the geocentric coordinate system based on the range-Doppler SAR positioning model, and converting the coordinates into latitude and longitude information, that is, the longitude Ω of the four corner points are respectively {Ω left_up Ω left_downn Ω right_up Ω right_down}, the latitudes Φ are respectively {Φ left_up Φ left_down Φ right_up Φ right_down}

[0140] Take the top left corner point Ω of the SAR image left_up and the latitude value Ω of the bottom left corner point left_down By comparing the scene pixels to determine whether they are consistent with the direction from north to south latitude, the latitude and longitude values ​​of all pixels in the SAR image can be calculated by constructing a kernel function based on the latitude and longitude of these four corner points.

[0141] First, calculate the row slope {k} of the longitude and latitude in the longitude and latitude grid. row_Φ k row_Ω} and column slope {k col_Φ k col_Ω} and the initial latitude and longitude values ​​{Ω start Φ start}, if Ω left_up >Ω left_down ,but:

[0142]

[0143] If Ω left_up <Ω left_down ,but:

[0144]

[0145] Further setting the intervals ΔΦ and ΔΩ for the latitude and longitude grid, the number of range points N and azimuth points M of the latitude and longitude grid are calculated based on the maximum and minimum longitude and latitude values ​​in the scene, thus obtaining the latitude and longitude values ​​corresponding to all pixels in the SAR image. The calculation formula is as follows:

[0146]

[0147] Step S800: Establish the index number mapping relationship between the coordinates of each pixel in the SAR image in the geocentric coordinate system and the geographic latitude and longitude grid.

[0148] Step S900: Based on the graphics processor, a kernel function is constructed to sequentially map the SAR image pixels in the imaging plane to the latitude and longitude grid, thus completing the geometric correction.

[0149] In steps S800 and S900, after dividing the latitude and longitude grid and obtaining the latitude and longitude values ​​of all pixels in the SAR image, a mapping relationship between the row and column numbers of SAR image pixels (i, j) and latitude and longitude grids (m, n) is established. Image pixels (i, j) are mapped one-to-one with threads in the thread block. This can be achieved by mapping the thread and block index to matrix coordinates to obtain the global memory offset of the thread block and thread index, and then mapping the matrix coordinates to the storage units in global memory. The correspondence between (i, j) and the index identifiers threadIdx.x, threadIdx.y, blockIdx.x, blockIdx.y, blockDim.x, and blockDim.y is as follows:

[0150]

[0151] Where threadIdx.x and threadIdx.y are the x-dimensional index and y-dimensional index of the thread in each thread block, respectively; blockIdx.x and blockIdx.y are the x-dimensional index and y-dimensional index of the thread block in the thread grid, respectively; and blockDim.x and blockDim.y are the x-dimensional size and y-dimensional size of the thread block, respectively.

[0152] The kernel functions for constructing the mapping relationship between the latitude and longitude values ​​of SAR image pixels (i, j) and the row and column numbers of latitude and longitude grids (m, n) are as follows:

[0153]

[0154] Finally, the SAR image is copied from the host to the device. The SAR image is divided into multiple sub-SAR images according to the GPU memory size. The grid size and thread block size are defined. The kernel function for geometric correction mapping is executed in the GPU to map all pixels in the sub-SAR image to the latitude and longitude grid space. The final SAR image result after radiometric and geometric correction is output.

[0155] like Figure 7 , Figure 8 As shown, Figure 7 This is the original SAR image of a certain scene. Figure 8 This image shows a SAR image after radiometric and geometric correction, obtained using the on-orbit radiometric and geometric correction method for spaceborne SAR images based on a parallel computing architecture provided in this invention. A comparison is made... Figure 7 and Figure 8 It can be seen that the SAR images after radiometric and geometric correction have relatively high image quality, which can ensure accurate positioning.

[0156] In summary, the on-orbit radiometric and geometric correction method for spaceborne SAR images based on a parallel computing architecture provided by this invention not only improves the quality of SAR images and the accurate positioning of SAR images, but also achieves efficiency and real-time processing of data processing tasks. It helps the on-orbit SAR processor to quickly complete radiometric and geometric correction tasks, while meeting the requirements of radiometric fidelity of SAR image products and accurate positioning of SAR images, and rapidly responding to task requirements while taking into account the high real-time requirements of tasks.

[0157] Figure 9 The diagram shows a structural schematic of an embodiment of an electronic device provided by the present invention. The specific embodiments of the present invention do not limit the specific implementation of the electronic device.

[0158] like Figure 9 As shown, the electronic device may include: a processor 502, a communications interface 504, a memory 506, and a communications bus 508.

[0159] The processor 502, communication interface 504, and memory 506 communicate with each other via communication bus 508. Communication interface 504 is used to communicate with other network elements, such as clients or other servers. Processor 502 executes program 510, specifically performing the relevant steps described above in the embodiment of the on-orbit radiometric and geometric correction method for spaceborne SAR images based on a parallel computing architecture.

[0160] Specifically, program 510 may include program code, which includes computer-executable instructions.

[0161] Processor 502 may be a central processing unit (CPU), a graphics processing unit (GPU), or an application-specific integrated circuit (ASIC), or one or more integrated circuits configured to implement embodiments of the present invention. The electronic device includes one or more processors, which may be processors of the same type, such as one or more CPUs; or they may be processors of different types, such as one or more CPUs and one or more ASICs.

[0162] Memory 506 is used to store program 510. Memory 506 may include high-speed RAM memory, and may also include non-volatile memory, such as at least one disk storage device.

[0163] Specifically, program 510 can be called by processor 502 to enable electronic equipment to execute the relevant steps in the above embodiment of the method for on-orbit radiometric and geometric correction of spaceborne SAR images based on parallel computing architecture.

[0164] Those skilled in the art will understand that Figure 9 The structure shown is for illustrative purposes only and does not limit the structure of the aforementioned device. For example, electronic devices may also include components that are more... Figure 9 The more or fewer components shown, or having the same Figure 9 The different configurations shown.

[0165] This invention also provides a computer-readable storage medium. The methods described above according to embodiments of the invention can be implemented in hardware or firmware, or implemented as computer code that can be recorded on a storage medium, or implemented as computer code downloaded via a network and originally stored on a remote storage medium or a non-transitory machine-readable storage medium and then stored on a local storage medium. Thus, the methods described herein can be processed by software stored on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. The storage medium can be a magnetic disk, optical disk, read-only memory, random access memory, flash memory, hard disk, or solid-state drive, etc.; further, the storage medium can also include combinations of the above types of memory. It is understood that computers, processors, microprocessor controllers, or programmable hardware include storage components capable of storing or receiving software or computer code, which, when accessed and executed by the computer, processor, or hardware, implements the methods shown in the above embodiments.

[0166] In the specific implementation of the above embodiments, the technical features can be combined in any non-contradictory way. For the sake of brevity, not all possible combinations of the above technical features are described. However, as long as the combination of these technical features is not contradictory, it should be considered to be within the scope of this specification.

[0167] The specific embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are detailed, they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention. Therefore, the scope of protection of this patent should be determined by the appended claims.

Claims

1. A method for on-orbit radiometric correction of spaceborne SAR images based on a parallel computing architecture, characterized in that, The method includes: Obtain the ephemeris parameters of the satellite in the geocentric coordinate system at various positions and times. The ephemeris parameters include at least one of the following: position vector, velocity vector, three-axis attitude angle, pulse repetition frequency, wave position sampling start time, signal pulse width, signal wavelength, antenna range aperture length, antenna azimuth aperture length, antenna beam elevation angle, antenna beam azimuth angle, number of range sampling points, number of azimuth sampling points, and signal sampling frequency. Calculate the slant distance between the satellite beam center pointing to the scene center point and the antenna phase center at each azimuth time, and the slant distance between each sampling point and the satellite at each azimuth time as well as the SAR echo distance at each azimuth time. Based on the slant distance between the satellite beam center pointing to the scene center point and the antenna phase center at each azimuth time, the slant distance between the SAR echo distance at each azimuth time and each sampling point and the satellite, and the ephemeris parameters, the downward angle between the satellite beam center pointing to each azimuth time and the line connecting the satellite and the nadir point is calculated. Based on the satellite beam center pointing at each of the aforementioned times and the downward angle between the satellite and the line connecting the nadir point, the antenna range-to-beam pattern is calculated and a radiation correction function is constructed. The calculation of the antenna range-direction beam pattern and the construction of the radiation correction function based on the satellite beam center pointing at the specified locations and the downward viewing angle between the satellite and the nadir point at each time point include: The difference between the downward angle of the satellite beam center and the line connecting the satellite and the nadir point at each location and the downward angle corresponding to the scene center is calculated using the formula shown in equation (4): (4) Based on the difference The antenna range-to-beam pattern is calculated using the formula shown in equation (5): (5) The radiation correction function is constructed based on the antenna range-direction beam pattern, and the calculation formula is shown in equation (6): (6) in, The downward angle between the satellite beam center pointing at each position and the line connecting the satellite and the nadir point; The downward perspective corresponding to the center of the scene; This represents the difference between the downward viewing angle corresponding to each sampling point and the downward viewing angle corresponding to the scene center at each azimuth time. This is a diagram showing the antenna range towards the beam direction. The distance to the aperture is the antenna distance; The wavelength of the signal; For radiation correction functions; The scaling factor is an adjustable factor after image normalization; n is a constant. ; The radiation distortion of SAR images is corrected in parallel based on the aforementioned radiation correction function.

2. The method according to claim 1, characterized in that, The parallel correction of radiometric distortion in SAR images based on the radiometric correction function includes: The SAR image is divided into multiple sub-SAR images; Based on the CUDA programming model, thread blocks are set up corresponding to the multiple sub-SAR images; The parallel processing of multiple thread blocks using a graphics processor completes the parallel correction of radiometric distortion in multiple sub-SAR images.

3. The method according to claim 1, characterized in that, The three-axis attitude angles include the ground roll angle, ground pitch angle, and ground yaw angle; the calculation of the slant range between the satellite beam center pointing to the scene center point and the antenna phase center at each azimuth time, and the slant range between the SAR echo range at each sampling point and the satellite at each azimuth time based on the ephemeris parameters, includes: Based on the three-axis attitude angles and the satellite's velocity and position vectors in the geocentric coordinate system, the number of nadir echo pulses and the SAR echo acquisition start time are calculated. The calculation formula is shown in equation (1): (1) Based on the start time of the echo acquisition and the number of echo pulses at the nadir point, the slant distance between the satellite at the beam center and the beam pointing point, and the slant distance between each sampling point and the satellite at each azimuth time SAR echo distance are calculated. The calculation formula is shown in Equation (2): (2) in, The speed of light; The slant distance between each sampling point and the satellite at each azimuth time is the SAR echo distance at each azimuth time. The slant distance between the point where the satellite beam center points to the center of the scene and the antenna phase center at each position and time. This is the start time of echo acquisition; This represents the number of echo pulses at the sub-satellite point. This is the start time of wave position sampling; The pulse repetition frequency; The signal pulse width; The signal sampling frequency; This represents the number of sampling points in the distance direction.

4. The method according to claim 3, characterized in that, The calculation of the downward angle between the satellite beam center pointing to the scene center point and the antenna phase center at each azimuth time, the slant distance between the SAR echo distance at each azimuth time and each sampling point and the satellite, and the ephemeris parameters, includes: The distance between the satellite and the nadir point is calculated based on the satellite's position vector and the Earth's average radius. Based on the slant distance between the satellite beam center pointing to the scene center point and the antenna phase center at each azimuth time, the slant distance between each sampling point and the satellite at each azimuth time, the distance between the satellite and the nadir point, and the average radius of the Earth, the downward angle between the satellite beam center pointing to the scene center point and the line connecting the satellite and the nadir point at each azimuth time is calculated; the calculation formula is shown in equation (3): (3) in, The downward angle between the satellite beam center pointing at each position and the line connecting the satellite and the nadir point; The distance between the satellite and the nadir point; This is the average radius of the Earth.

5. An electronic device, characterized in that, Also includes: The system includes a memory, a processor, a communication interface, and a communication bus, wherein the processor, the memory, and the communication interface communicate with each other via the communication bus. The memory is used to store at least one executable instruction that causes the processor to execute the on-orbit radiometric correction method for spaceborne SAR images based on a parallel computing architecture as described in any one of claims 1 to 4.

6. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that cause a processor to execute the on-orbit radiometric correction method for spaceborne SAR images based on a parallel computing architecture as described in any one of claims 1-4.

Images